Technology and Integration Roadmap for Optoelectronic Memristor

Advanced Materials, Journal Year: 2023, Volume and Issue: 36(9)

Published: Sept. 23, 2023

Optoelectronic memristors (OMs) have emerged as a promising optoelectronic Neuromorphic computing paradigm, opening up new opportunities for neurosynaptic devices and systems. These OMs possess range of desirable features including minimal crosstalk, high bandwidth, low power consumption, zero latency, the ability to replicate crucial neurological functions such vision optical memory. By incorporating large-scale parallel synaptic structures, are anticipated greatly enhance high-performance low-power in-memory computing, effectively overcoming limitations von Neumann bottleneck. However, progress in this field necessitates comprehensive understanding suitable structures techniques integrating low-dimensional materials into integrated circuit platforms. This review aims offer overview fundamental performance, mechanisms, design applications, integration roadmap memristors. establishing connections between materials, multilayer memristor units, monolithic circuits, seeks provide insights emerging technologies future prospects that expected drive innovation widespread adoption near future.

Language: Английский

---

Monolithic 2D Perovskites Enabled Artificial Photonic Synapses for Neuromorphic Vision Sensors

Advanced Materials, Journal Year: 2024, Volume and Issue: 36(18)

Published: Jan. 26, 2024

Abstract Neuromorphic visual sensors (NVS) based on photonic synapses hold a significant promise to emulate the human system. However, current rely exquisite engineering of complex heterogeneous interface realize learning and memory functions, resulting in high fabrication cost, reduced reliability, energy consumption uncompact architecture, severely limiting up‐scaled manufacture, on‐chip integration. Here photo‐memory fundamental ion‐exciton coupling is innovated simplify synaptic structure minimize consumption. Due intrinsic organic/inorganic within crystal, photodetector monolithic 2D perovskite exhibits persistent photocurrent lasting about 90 s, enabling versatile functions. The electrical power per event estimated be≈1.45 × 10 −16 J, one order magnitude lower than that natural biological Proof‐of‐concept image preprocessing using neuromorphic vision enabled by synapse demonstrates 4 times enhancement classification accuracy. Furthermore, getting rid artificial neural network, an expectation‐based thresholding model put forward mimic system for facial recognition. This conceptual device unveils new mechanism structure, promising transformation NVS fostering emergence next generation networks.

Language: Английский

---

Inorganic Halide Perovskite Nanowires/Conjugated Polymer Heterojunction-Based Optoelectronic Synaptic Transistors for Dynamic Machine Vision

Nano Letters, Journal Year: 2024, Volume and Issue: 24(14), P. 4132 - 4140

Published: March 27, 2024

Inspired by the retina, artificial optoelectronic synapses have groundbreaking potential for machine vision. The field-effect transistor is a crucial platform that highly sensitive to external stimuli and can modulate conductivity. On basis of decent optical absorption, perovskite materials been widely employed constructing synaptic transistors. However, reported transistors focus on static processing independent at different moments, while natural visual information consists temporal signals. Here, we report CsPbBrI2 nanowire-based study dynamic responses time-varying first time. Moreover, behavior, hardware system with an accuracy 85% built trajectory moving objects. This work offers new way develop construction vision systems.

Language: Английский

---

Low-Power Perovskite Neuromorphic Synapse with Enhanced Photon Efficiency for Directional Motion Perception

ACS Applied Materials & Interfaces, Journal Year: 2024, Volume and Issue: 16(17), P. 22303 - 22311

Published: April 16, 2024

The advancement of artificial intelligent vision systems heavily relies on the development fast and accurate optical imaging detection, identification, tracking. Framed by restricted response speeds low computational efficiency, traditional optoelectronic information devices are facing challenges in real-time tasks their ability to efficiently process complex visual data. To address limitations current devices, this study introduces a novel photomemristor utilizing halide perovskite thin films. fabrication involves adjusting iodide proportion enhance quality films minimize dark current. exhibits high external quantum efficiency over 85%, which leads energy consumption 0.6 nJ. spike timing-dependent plasticity characteristics device leveraged construct spiking neural network achieve 99.1% accuracy rate directional perception for moving objects. notable results offer promising hardware solution efficient optoneuromorphic edge computing applications.

Language: Английский

---

Tunable optoelectronic response in van der Waals heterojunction transistors for artificial visual recognition

Applied Physics Letters, Journal Year: 2024, Volume and Issue: 124(12)

Published: March 18, 2024

Optoelectronic synaptic transistors are advantageous in in-memory light sensing for artificial neural networks. Herein, optoelectronic junction field-effect (JFETs) based on a Ga2O3/MoS2 heterojunction fabricated. The devices exhibit robust electrical performances, including high on/off ratio of 108, low subthreshold swing 69 mV dec−1, and output current 3.4 μA μm−1. An inverter NAND gate constructed the dual-gated configuration, with showing voltage gain 28 near-ideal noise margin 90.4%. Additionally, demonstrate outstanding performances benefiting from strong light–matter interactions MoS2. Typical plasticities, short-term plasticity, long-term spiking-rate-dependent simulated by applying pulses. Furthermore, metaplastic excitatory postsynaptic current, facilitation potentiation transition to depression also readily demonstrated. network, which neurons interconnected through our proposed transistors, achieves accuracy 89.8% recognizing handwritten digits. This work provides insight into design an synapse JFETs.

Language: Английский

---

Perovskite micro-/nanoarchitecture for photonic applications

Matter, Journal Year: 2023, Volume and Issue: 6(10), P. 3165 - 3219

Published: June 25, 2023

Language: Английский

---

Research progress and applications of optoelectronic synaptic devices based on 2D materials

Brain‐X, Journal Year: 2024, Volume and Issue: 2(3)

Published: Sept. 1, 2024

Abstract In the natural world, human brain is most powerful information processor, using a highly parallel, efficient, fault‐tolerant, and reconfigurable neural network. Taking inspiration from this impressive architecture, optoelectronic synaptic devices have gained considerable attention for their ability to process retain data simultaneously, making them essential components in upcoming era of neuromorphic computing systems. recent years, significant progress has been made development two‐dimensional (2D) material heterostructures. This review focuses on use 2D materials creating devices. It discusses utilizing heterostructures these examines potential different areas such as image recognition, wearable electronics, logical operations, Heterostructures with provide wide range possibilities electronic band structures can be easily tailored achieve effective optical electrical modulation. Optoelectronic based simultaneously exhibit two functionalities: detection memory. Furthermore, strong interatomic bonding within layers possess thickness only one atomic layer, giving exceptional flexibility, transparency, mechanical strength. By solution processing ultra‐thin profile, manufacturing three‐terminal synapses becomes cost‐effective, simplifying integration processes.

Language: Английский

---

Neuromorphic optoelectronic devices based on metal halide perovskite

Materials Today Electronics, Journal Year: 2024, Volume and Issue: 8, P. 100099 - 100099

Published: April 22, 2024

Neuromorphic electronics has received increased attention for their application in brain-inspired computing and artificial sensorimotor nerves. Metal halide perovskite (MHP) been proved to be a candidate material use optoelectronic neuromorphic devices. Herein, we review on the recent research progress of MHP materials, with focus applications optoelectronics. First, materials that are used devices especially applications, sequence all-inorganic, organic-inorganic hybrid lead-free materials. Then, summarize design fabrication two-terminal (2-T) three-terminal (3-T) synaptic devices, including working mechanisms as operated by electrical light inputs, relationship between properties composition structure functional layers. Finally, these pattern recognition, bionic vision, sensing modulation, experience associative learning, logic high-pass filtering. This aims could potentially inspire future field electronics.

Language: Английский

---

Thermally and Mechanically Stable Perovskite Artificial Synapse as Tuned by Phase Engineering for Efferent Neuromuscular Control

Nano Letters, Journal Year: 2024, Volume and Issue: 24(30), P. 9311 - 9321

Published: July 18, 2024

The doping of perovskites with mixed cations and halides is an effective strategy to optimize phase stability. In this study, we introduce a cubic black perovskite CsyFA(1–y)Pb(BrxI(1–x))3 artificial synapse, using engineering by adjusting the cesium-bromide content. Low-bromine are suitable improve electric pulse excitation sensitivity stability device. Specifically, low-bromine low-cesium (x = 0.15, y 0.22) annealed at 373 K allows device maintain logic response even after 1000 mechanical flex/flat cycles. also shows good thermal up temperatures 333 K. We have demonstrated reflex-arc behavior MCMHP synaptic units, capable making sensory warnings high frequency. This compositionally engineered, dual-mixed provides significant potential for perceptual soft neurorobotic systems prostheses.

Language: Английский

---

Carbon-based memristors for resistive random access memory and neuromorphic applications

Chip, Journal Year: 2024, Volume and Issue: 3(2), P. 100086 - 100086

Published: Feb. 2, 2024

As a typical representative of nanomaterials, carbon nanomaterials have attracted widespread attention in the construction electronic devices due to their unique physical and chemical properties, multi-dimensionality, multi-hybridization methods excellent properties. Especially recent years, memristors based on flourished field building non-volatile memory neuromorphic applications. In this paper, preparation structural characteristics different dimensions are systematically reviewed. Then, working mechanism discussed depth. Finally, potential applications carbon-based logic operations, neural network construction, artificial vision systems, tactile multimodal perception systems introduced. We believe paper will provide guidance for future development high-quality information storage, high-performance applications, high-sensitivity bionic sensing memristors.

Language: Английский

---